VCCV: conservative transcriptomic corroboration for measurement prioritization of computational drug–target hypotheses

Dingkui Kang

doi:10.1093/bioinformatics/btag694

VCCV: conservative transcriptomic corroboration for measurement prioritization of computational drug–target hypotheses | ResearchHub

Abstract

Abstract Motivation Computational drug-target interaction (DTI) models nominate plausible binders but cannot determine which candidate best accounts for an observed cellular response. Perturbational transcriptomics offers orthogonal mechanistic evidence, yet pharmacology-to-genetics mismatch, incomplete reference coverage, and non-specific stress programs make simple signature matching unreliable. This motivates a principled integration layer that corroborates hypotheses conservatively, abstains under global mismatch, and prioritizes informative follow-up measurements when the evidence remains ambiguous. Results We present Virtual-to-Cellular Corroboration for Validation (VCCV), a model-agnostic posterior-triage layer for pre-trained DTI models. VCCV updates calibrated DTI working weights with context-aligned perturbational evidence using a near-identity affine map and exact covariance transport. Within the stated Gaussian class, exact transport is the unique uncertainty update that preserves posterior-odds comparisons under invertible affine changes of measurement coordinates. VCCV also introduces an empirical warning branch for abstention and converts residual ambiguity into compact follow-up gene panels, using a submodular objective for deep near-ties. Each query is assigned one of three actionable states: a target-resolved hypothesis, an abstention, or a prioritized panel. Across five DTI models, VCCV improved discrimination (paired ROC-AUC gains 0.021–0.037) and reduced negative log-likelihood in every case. Additional evaluations showed improved ranking on the same-cell-supported endpoint, warning-score discrimination of strong-response profiles (ROC-AUC 0.876), and better recovery of full-coordinate leading hypotheses by selected panels than by random panels. Across these retrospective kinase-focused evaluations, VCCV provides a principled bridge from computational nomination to conservative, measurement-directed cellular corroboration. Availability and implementation Source code of VCCV is publicly available at https://github.com/bio-ai-source/VCCV.

External reference records

Polypharmacology: challenges and opportunities in drug discovery: miniperspective

10.1021/jm5006463 · 2014

Drug target prediction through deep learning functional representation of gene signatures

10.1038/s41467-024-46089-y · 2024

scGPT: toward building a foundation model for single-cell multi-omics using generative AI

10.1038/s41592-024-02201-0 · 2024

Comprehensive analysis of kinase inhibitor selectivity

10.1038/nbt.1990 · 2011

MolTrans: molecular interaction transformer for drug–target interaction prediction

10.1093/bioinformatics/btaa880 · 2021

Target deconvolution from phenotype-based drug discovery by using chemical proteomics approaches

10.1016/j.bbapap.2018.08.002 · 2019

The Connectivity Map: using gene-expression signatures to connect small molecules, genes, and disease

10.1126/science.1132939 · 2006

Predicting emerging drug interactions using GNNs

10.1038/s43588-023-00555-7 · 2023

Specific gene module pair-based target identification and drug discovery

10.3389/fphar.2022.1089217 · 2023

SP-DTI: subpocket-informed transformer for drug–target interaction prediction

10.1093/bioinformatics/btaf011 · 2025

DTIAM: a unified framework for predicting drug–target interactions, binding affinities and drug mechanisms

10.1038/s41467-025-57828-0 · 2025

An analysis of approximations for maximizing submodular set functions, I

10.1007/bf01588971 · 1978

GraphDTA: predicting drug–target binding affinity with graph neural networks

10.1093/bioinformatics/btaa921 · 2021

Improving generalizability of drug–target binding prediction by pre-trained multi-view molecular representations

10.1093/bioinformatics/btaf002 · 2025

DeepDTA: deep drug–target binding affinity prediction

10.1093/bioinformatics/bty593 · 2018

10.7551/mitpress/1113.003.0008

10.7551/mitpress/1113.003.0008 · 2000

Predicting transcriptional responses to novel chemical perturbations using deep generative model for drug discovery

10.1038/s41467-024-53457-1 · 2024

Target identification and mechanism of action in chemical biology and drug discovery

10.1038/nchembio.1199 · 2013

DeepDTAGen: a multitask deep learning framework for drug–target affinity prediction and target-aware drugs generation

10.1038/s41467-025-59917-6 · 2025

LINCS Data Portal 2.0: next generation access point for perturbation-response signatures

10.1093/nar/gkz1023 · 2020

A next generation connectivity map: l 1000 platform and the first 1,000,000 profiles

10.1016/j.cell.2017.10.049 · 2017

How were new medicines discovered?

10.1038/nrd3480 · 2011

Making sense of large-scale kinase inhibitor bioactivity data sets: a comparative and integrative analysis

10.1021/ci400709d · 2014

Transfer learning enables predictions in network biology

10.1038/s41586-023-06139-9 · 2023

MLG2Net: molecular global graph network for drug response prediction in lung cancer cell lines

10.1007/s10916-025-02182-3 · 2025

PerturbNet predicts single-cell responses to unseen chemical and genetic perturbations

10.1038/s44320-025-00131-3 · 2025

GS-DTI: a graph-structure-aware framework leveraging large language models for drug–target interaction prediction

10.1093/bioinformatics/btaf445 · 2025

Evidential deep learning-based drug–target interaction prediction

10.1038/s41467-025-62235-6 · 2025

Additional external references

No additional external references are available.

VCCV: conservative transcriptomic corroboration for measurement prioritization of computational drug–target hypotheses

Referenced works

Cited by

External reference records

Additional external references